The goal of this clinical trial is to learn if the VELYS Robotic-Assisted Solution (VRAS) device helps to improve outcomes from total knee replacement, when compared to regular non-robotic-assisted total knee replacement in the treatment of osteoarthritis. It will also aim to review the longer-term safety and efficiency effects of using this device.

The main questions it aims to answer are:

1. Does VRAS device-assisted total knee replacement improve patient outcomes (such as their function, satisfaction, and pain) when compared to non-robotic-assisted total knee replacement?
2. Does VRAS device-assisted total knee replacement improve clinical outcomes (such as revision risk, complications, and hospital length of stay) when compared to non-robotic-assisted total knee replacement?
3. Does VRAS device-assisted total knee replacement improve functional outcomes (including biomechanical motion analysis) when compared to non-robotic-assisted total knee replacement?
4. What are the economic and efficiency effects of VRAS device-assisted total knee replacement compared to non-robotic-assisted total knee replacement?
5. What are the human factor and learning curve effects of introducing VRAS device-assisted total knee replacement into healthcare teams?
6. Are there any differences in safety and adverse event incidence between VRAS device-assisted total knee replacement and non-robotic-assisted total knee replacement?

Participants will:

• Be randomised to receive a total knee replacement using either regular non-robotic-assisted methods, or with the use of the VRAS device.
• Be assessed pre-operatively at the clinic, and then at 6 weeks, 6 months, and 12 months following date of operation. At these appointments patients will answer questionnaires, receive an x-ray (pre-operatively, and at 6 months post-operatively), and a subgroup will be reviewed in the motion analysis laboratory (pre-operatively, and at 6 months post-operatively).
• Have their longer-term outcomes including any revision operations monitored out to 10 years through national registry data linkage (no actual follow-up for patients after 12 months).

This multicentre, pragmatic, parallel-group, blinded RCT will compare VELYS robot-assisted functional alignment with conventional total knee arthroplasty. The trial will employ a clearly defined, reproducible surgical protocol that will incorporate evaluation of the study cohort using: established clinical and patient-reported outcome measures (PROMs); gold-standard three-dimensional (3D) motion analysis to objectively evaluate functional outcomes; detailed examination of intraoperative knee kinematics, pre-/post-operative hip-knee-ankle radiographic data, and healthcare system effects including economic, efficiency, and human factor measures.

Objectives

The primary objective is to compare knee-specific health outcomes in patients undergoing TKA performed with VELYS robot-assisted (raTKA) using functional alignment versus TKA performed manually (mTKA) using conventional instrumentation. The primary outcome measure will be differences between the two intervention groups (using an intention-to-treat approach) with regard to change in Forgotten Joint Score (FJS) preoperatively versus 6 months post-surgery. The secondary objectives are to:

1. Compare PROMs (FJS, EQ-5D-5L, Oxford Knee Score [OKS], satisfaction [numeric rating scale; NRS], 4-part Likert scale], pain [101-point scale VAS]) at 6 weeks, 6 months, and 12 months;

2. Correlate these outcomes with radiographic alignment and intra-operative data;

3. Evaluate biomechanical outcomes using motion analysis laboratory assessment;

4. Functional and activity-related outcomes;

5. Evaluate the economic and efficiency effects of imageless raTKA relative to mTKA;

6. Evaluate the human factor effects of introducing a new RAS into healthcare teams;

7. Assess surgical safety and the incidence of adverse events in both groups.

   Null hypothesis The null hypothesis is that functionally aligned raTKA using the VELYS imageless system does not result in a greater early improvement in knee-specific outcomes, as measured by change in Forgotten Joint Score at 6 months, compared with conventional mTKA.

   Study design The study is designed as a multicentre, pragmatic, parallel-group blinded RCT. There are nested modular study components, including: an internal pilot to determine safety and practicability; a motion analysis evaluation, and a task load, efficiency, and clinician perspective evaluation. The lead centre for the study will be the Golden Jubilee University National Hospital (GJUNH). Additional centres will be based in the United Kingdom and Irish regions. All participating surgeons are high-volume arthroplasty surgeons (performing >100 knee arthroplasties annually) and have extensive experience with robot-assisted surgery.

   Surgeons and training All participating surgeons are high-volume arthroplasty surgeons (performing >100 knee arthroplasties annually) and have extensive experience with robot-assisted surgery.

   To minimise any learning curve effects and variability between centres, all surgeons will have completed manufacturer-delivered training and credentialling using both the manual and robotic systems, and a minimum of 20 Attune TKA non-trial cases will have been completed by all surgeons (comprising at least 10 manual Attune and 10 VELYS-assisted cases) before they recruit trial participants. A subgroup analysis of TKA performed early versus late on the clinical experience will be performed as part of the interim pilot study to evaluate for differences in outcomes and adverse events, and a sensitivity analysis will be performed if indicated.

   Sample size The sample size calculation is based on detecting a minimum clinically important difference (MCID) of 7 points in the FJS at six months post-operatively. This is informed by Alton et al (2025), and a standard deviation of 24 points, as reported in Clement et al., 2021. Using a one-sided alpha of 0.05 and a power of 80%, it was calculated that 147 participants per group would be required. To account for an anticipated 15% loss to follow-up, the final recruitment target is 173 participants per group, yielding a total study population of 346 participants.

   Outcome data from the internal pilot (the first 100 study participants who have completed their 6-month follow-up) will be analysed. A blinded pooled estimate of the outcome variance will be computed and used in a pre-specified sample size re-estimation algorithm. The analysis team will remain blinded to treatment allocation; any sample size adjustments will be made according to the pre-specified rules to preserve type I error.

   For the motion analysis sub-study, a sample size of 52 participants (26 in the raTKA and mTKA groups respectively) has been determined to provide sufficient power to detect expected differences in peak medial knee contact force compared with matched healthy controls.

   Recruitment Participants will be identified through routine NHS orthopaedic outpatient clinics at the participating centres, in accordance with the eligibility criteria by the clinical team. Potentially eligible patients will be highlighted by the research team who will provide patient information relating to the study, with adequate time for consideration of the relevant materials and discussion with the clinical team before consent is sought. Participants will be randomised on a 1:1 basis to receive either: mTKA or raTKA. A separate consent form will be provided for the MAL sub-study.

   Recruitment is planned to take place over 36 months, with each participant followed for 12 months post-operatively. Due to the use of the Attune Medial Stabilized (MS) polyethylene insert, long-term observation will be conducted for a minimum of 10 years to evaluate survivorship and cross-validated with national registry data.

   Randomisation Participants will be randomised in a 1:1 ratio to receive either manual TKA or VELYS robot-assisted functional alignment TKA. Randomisation will be performed using a secure web-based system (Research Electronic Data Capture [REDCap], Vanderbilt University, USA) incorporating computer-generated permuted block randomisation, stratified by recruitment centre. Block sizes will be randomly varied and concealed to ensure allocation concealment. Following written informed consent and collection of baseline data, site research staff will enroll and randomise participants.

   All participants will receive standard perioperative and postoperative care according to local institutional practice, including anaesthesia, thromboprophylaxis, analgesia, mobilisation, physiotherapy, and routine follow-up. Concomitant care considered part of usual care is permitted. No additional trial-specific co-interventions are mandated other than the allocated surgical approach and study assessments. Use of alternative robotic systems, non-protocolised experimental surgical technologies, or non-standard implants or constraint not anticipated by the protocol will not be permitted unless clinically required in the interests of patient safety. Any such deviations and the reasons for them will be documented. Participants will continue to receive standard NHS care following their involvement in the study. No additional or special post-trial care is required.

   Blinding The study is a randomised controlled trial with patient and outcome assessor blinding. It is not possible to blind surgeons given the nature of the interventions. Non-surgical clinical staff (e.g. ward nurses and allied health professionals) will be blinded. Femoral and tibial tracker pins will be placed within the surgical incision, avoiding the need for accessory incisions in raTKA cases or sham incisions to aid blinding in mTKA cases. Documentation will be redacted appropriately to prevent inadvertent unblinding of clinicians. PROMs will be completed independently by participants. Radiographic assessments and biomechanical analyses will be performed by assessors who are blinded to treatment allocation. All study data will be anonymised and the study arm concealed before the conduct of data analysis. No unblinding procedures are planned.

   Routine unblinding is not planned. Unblinding will be permitted only where knowledge of treatment allocation is considered necessary for urgent clinical management or participant safety. Any request for unblinding should be made by the treating clinician to the Chief Investigator or delegated medically qualified investigator. The reason for unblinding, the date, the person authorising it, and the person informed will be documented in the trial records. The Sponsor will be informed, and the event will be reviewed by the Trial Steering Committee. Wherever possible, unblinding of one participant will not result in unblinding of outcome assessors or the wider trial team.

   Withdrawal of subjects

   Participants have the right to withdraw from the trial at any point and for any reason, without providing an explanation. Withdrawal from the study will not affect their routine clinical care. The Principal Investigators (or delegate) may also withdraw participants from the study intervention if it is deemed in their best interest, including but not limited to:

   • Development of a medical condition making continuation inappropriate;
   • Adverse Events (AEs) or Serious Adverse Events (SAEs) that necessitate discontinuation;
   • Significant protocol violations that compromise study integrity or participant safety;
   • Participant request or non-compliance with study procedures.

   Data collected up to the point of withdrawal will be retained and included in analyses, unless the participant specifically requests data removal in line with GDPR provisions. Participants who withdraw before randomisation or before any study procedures will be replaced to ensure an adequate sample size. Participants who withdraw after randomisation will not be routinely replaced, unless withdrawal rates threaten study power, in which case replacement will be considered. If withdrawal occurs due to an AE or SAE, appropriate medical care will be provided and relevant safety follow-up conducted until the event has resolved or stabilised. All such cases will be reported in accordance with regulatory requirements. Participants will be informed of these arrangements during the consent process, and all efforts will be made to minimise the burden of continued safety follow-up for those who withdraw from the intervention arm of the study.

   Surgical technique A thigh tourniquet will be used for the full case until closure of the arthrotomy, and a medial parapatellar approach to the knee will be used in all cases. A cemented Attune prosthesis with medial stabilised (MS) polyethylene insert will be implanted in all patients. Perioperative care will be delivered in a standardised fashion in accordance with safe established practice and will not vary according to treatment arm.

   Manual TKA Participants randomised to this group will undergo manual TKA performed with conventional jig-based instrumentation using cemented Attune Knee System components (Johnson & Johnson MedTech) and a medial stabilised (MS) polyethylene insert. The aim will be to correct deformity towards a neutral limb alignment and achieve a TKA that is stable throughout the flexion-extension arc.

   Extramedullary referencing will be used to align the proximal tibial resection perpendicular to the axis of the tibia. Intramedullary referencing will be used to align the distal femoral resection based on pre-operative planning and intra-operative assessment. Femoral axial alignment will be set initially at 3 degrees of external rotation relative to the posterior condylar axis, typically perpendicular to Whiteside's line, and adjusted based on the flexion gap and patellar tracking. Measured bone resections will be made using the standard cutting guides for the Attune TKA system.

   VELYS robot-assisted functional alignment TKA Participants allocated to the raTKA group will undergo TKA using the VELYS system following a functional alignment (FA) strategy that uses technology-assisted intraoperative evaluation to restore ligament balance and stability throughout the flexion-extension arc while reconstructing the native kinematic axes of the pre-arthritic knee. This is achieved by precise assessment of the soft tissue envelope using the VELYS system with subsequent adjustments to the planned bone resections to achieve an optimally sized, aligned, and balanced TKA using minimal bone resections to accommodate the thinnest necessary polyethylene insert and minimising the need for soft tissue releases. Bone resections and implant positioning will be guided by the VELYS Robot-assisted Solution within safe target ranges: tibial resection 0° to 7° varus; overall limb alignment (HKA) 173°-183° (i.e. HKA between 7° varus and 3° valgus). The femoral and tibial robotic arrays will be placed intra-incisional according to the standard VLEYS technique to avoid the need for additional incisions.

   It is possible to achieve FA either from a mechanically-aligned pre-operative plan, where bone resections are planned to produce neutral alignment in the coronal plane, or a kinematically-aligned pre-operative plan, where anatomical bone resections are planned to resurface the native anatomy. The latter accounts for bone loss, and thus restores the constitutional [pre-arthritic] limb alignment and joint line obliquity. FA is then achieved by technology-assisted intraoperative assessment of the soft tissues and subsequent adjustments made to the planned resections to achieve a balanced knee. Soft tissue release is only utilised as a last resort if satisfactory alignment or balance is not achieved by modifying the bone resections alone. For consistency in this study, all raTKA patients will be treated according to a KA-based FA philosophy. VELYS-acquired intra-operative data will be captured for correlation with post-operative outcomes.

   Technical consistency The Surgical Lead will chair a surgeon's group to review the surgical and perioperative techniques to ensure consistency of practice across all participating sites. Any deviations from the defined surgical protocol will be documented and will require prior approval by the CI and site PIs, Trial Steering Committee (TSC) and Research Ethics Committee (REC). Adherence to the intervention protocol will be monitored via surgical records and VELYS system logs.

   Outcomes The primary outcome will be the change in Forgotten Joint Score (FJS) at two timepoints, preoperatively versus 6 months post-surgery. Secondary outcome measures will include PROMs; clinical outcomes; functional, physiological, and technology-enabled activity measures; health service and economic outcomes; technical surgical/radiographic measures, and biomechanical outcomes using motion analysis laboratory assessment. A dedicated motion analysis laboratory (MAL) study will be conducted as a nested sub-study.

   Subgroup analyses Predefined subgroup analyses will explore potential differences in outcomes according to: age; sex; BMI; socioeconomic deprivation level; primary compartment involved (medial, lateral, or patellofemoral); CPAK group and restoration; VELYS-acquired intra-operative data (for raTKA group).

   These subgroup analyses will extend the primary models by including interaction terms and will be considered exploratory. A generalised linear model will be applied, incorporating allocation group, age, site, surgeon, gender, BMI (≥35 kg/m2), and the primary compartment affected (medial, lateral, or patellofemoral) as covariates. Both fixed- and random-effects modelling will be employed to assess robustness. Sensitivity analyses will explore model assumptions and include both modelling approaches. Other subgroup analyses may be pre-specified as appropriate.

   Data collection Study data will be collected using a bespoke portal on REDCap (Research Electronic Data Capture, Vanderbilt University, USA) at predefined time points, aligned with the clinical care pathway. An option for in-person, postal, and telephone collection will also be available to participants at each time point.

   At the pre-operative baseline visit, participants will complete the PROMs questionnaires (FJS, OKS, EQ-5D-5L, satisfaction scores, and pain score). Pre-operative bilateral hip-knee-ankle (HKA) radiographs will be obtained to assess coronal alignment, joint line obliquity, and enable CPAK classification. For participants enrolled in the motion analysis laboratory (MAL) sub-study, a baseline assessment will be conducted during the same hospital visit.

   During surgery, data will be collected relating to surgical workflow and efficiency (including case time, tourniquet time, operating room preparation and turnaround times). For the raTKA group, VELYS-acquired intra-operative data will be collected. Data from inpatient medical records will be collected. Routine postoperative radiographs will be taken in line with standard clinical practice.

   At six weeks post-operatively, participants will complete the PROMs questionnaires. In addition, participants will complete a health service resource use questionnaire, which will capture clinical service contacts (including frequency and reason) and analgesic use during the follow-up period.

   At six months post-operatively, participants will complete the FJS (primary outcome), and secondary PROMs questionnaires, and post-operative bilateral HKA radiographs will be obtained to assess implant positioning and limb alignment. A further health service resource use questionnaire will also be completed. MAL sub-study participants will undergo a repeat assessment.

   At 12 months post-operatively, participants will complete the full set of PROMs, along with a final health service resource use questionnaire. The final phase of data analysis and dissemination will be conducted in year 5 of the study. The trial protocol is aligned with SPIRIT 2025 guidelines. The study conduct will be reported according to CONSORT 2025 guidelines.

   Data management All study data will be recorded using case report forms (CRFs) and entered into the secure, password-protected REDCap platform. Access will be restricted to authorised members of the trial team and sponsor representatives. Personal identifiers will be replaced with coded participant IDs to ensure data confidentiality. De-identified data will be used for all statistical analyses. Data will be managed in accordance with the UK Caldicott principles, and data will be retained in excess of the minimum eight years in accordance with sponsor policy and UK regulatory requirements. Data quality will be safeguarded through range checks, source data verification, and ongoing data monitoring by the trial team. Audits may be conducted by the Sponsor in line with routine NHS research governance.

   Statistical analysis All analyses will adhere to CONSORT reporting standards. A detailed statistical analysis plan will be finalised with the Trial Steering Committee before any formal analyses are undertaken. Baseline characteristics will be summarised to assess comparability between treatment groups. Simple descriptive statistics will follow standard conventions, including means and standard deviations or medians and interquartile ranges as appropriate for continuous variables, and counts and percentages for categorical data.

   The primary analysis will examine differences between the two intervention groups (using an intention-to-treat approach) with regard to their change in FJS (preoperatively versus 6 months post-surgery). A Student's independent t-test will be used to compare mean scores between groups. Assumptions of normality and equality of variances will be assessed using diagnostic plots and Levene's test. If variances are unequal, the Welch's t-test will be applied. The results will be reported as mean differences with 95% confidence intervals and p-values.

   Secondary outcomes will be analysed similarly, appropriate to the type and distribution of the data. Categorical outcomes, such as patient satisfaction and complication rates, will be compared between groups using the chi-square test. If any expected cell count is less than five, Fisher's exact test will be applied. Effect sizes will be reported alongside these comparisons using risk ratios or odds ratios with 95% confidence intervals to provide a measure of the magnitude and precision of differences between groups.

   Internal pilot This study will include an internal pilot phase, during which feasibility parameters will be assessed in an interim analysis of the first 100 participants, followed up to 6 months. Data from this phase will contribute to the definitive trial analysis. Pre-specified progression criteria will guide continuation: recruitment ("Go" ≥75% of target rate; "Amend" 50-74%; "Stop" <50%), retention at the primary endpoint ("Go" ≥90%; "Amend" 75-89%; "Stop" <75%), intervention delivery fidelity ("Go" ≥85%; "Amend" 60-84%; "Stop" <60%), and completeness of primary outcome data ("Go" ≥90%; "Amend" 70-89%; "Stop" <70%). Safety signals or unexpected serious adverse events will prompt immediate review by the Trial Steering Committee. At the end of the internal pilot, the Trial Steering Committee will review feasibility data and determine whether to proceed without modification, continue with protocol amendments, or stop the trial.

   Motion analysis laboratory (MAL) evaluation A subset of 26 participants in the VELYS-assisted group and 26 participants in the manual group will undergo gold-standard 3D motion analysis (Vicon, Oxford, UK) to provide objective biomechanical (kinematic, kinetic and electromyographic) data which will be compared between groups and with a reference population of 52 age- and sex-matched healthy (native knee) controls using a 1:1:2 comparison ratio.

   The primary outcome measure will be knee adduction moment (KAM), which is a validated biomechanical indicator of medial compartment loading, with elevated KAM values associated with increased medial knee contact forces and implant wear. Systematic review evidence demonstrates that manual TKA procedures consistently reduce peak KAM. However, not many studies have systematically evaluated KAM patterns following raTKA or compared these outcomes to healthy individuals, representing a significant gap in understanding the biomechanical benefits of this advanced surgical technique. Secondary outcomes will include:

   • Spatiotemporal gait parameters (gait speed, step length);
   • Kinematic parameters (knee range of motion in sagittal and frontal planes);
   • Kinetic parameters, including peak hip flexion and adduction moments, peak knee flexion moment, knee adduction moment impulse, peak ankle dorsiflexion moment, and plantar loading symmetry;
   • Joint power parameters, including hip, knee, and ankle power;
   • Electromyographical assessment.

   A sample size of 52 participants (26 in the raTKA and mTKA groups, respectively) has been determined to provide sufficient power to detect expected differences in medial knee contact force compared with matched healthy controls. The sample size calculation is based on peak knee adduction moment (KAM). Peak KAM for healthy controls is described as being 0.46 ± 0.13 Nm/kg, and for prosthetic knees at six months after mTKA, it is 0.61 ± 0.13 Nm/kg. Only one study has investigated peak KAM in patients who have undergone raTKA, demonstrating values of 0.49 ± 0.31 Nm/kg at six months post-operatively. A power calculation was performed (SPSS Sample Power) and demonstrated that with an independent samples t-test, a sample of 26 subjects in each group (mTKA and raTKA) provided adequate power to detect a significant difference in peak KAM (large effect size [0.80], α = 0.05). Complementary MAL and technology-enabled assessments of real-world function, additional timepoints, and a larger sample may be added according to the availability of resources.

   The motion analysis laboratory is based at the GJUNH, and subjects for the MAL study will be recruited from the GJUNH cohort of trial subjects. Patients will be invited to participate in the MAL study at the point of recruitment to the trial until the full sample (26 patients in each experimental arm) is complete. Recruitment to the MAL study will require separate informed consent due to the additional requirements for in-person attendance and testing pre-operatively and at 6 months post-operatively. Patients with a BMI of >40 kg/m2 will be excluded from the MAL study due to potential line-of-sight issues and greater soft tissue artefacts associated with anatomical placement of the retro-reflective markers and their subsequent movement relative to underlying bone during trials.

   Health economic analysis A prospective economic evaluation will be conducted from the perspective of the NHS and personal social services. Participants' healthcare contacts related to knee arthroplasty will be collected at all follow-up points, along with time lost from paid or unpaid work. Differences in surgical resource use, including operating time and facility utilisation, will be costed using the latest published national reference costs, adjusted to a common year.

   Health-related quality of life will be measured using the EQ-5D-5L questionnaire, with scores converted to utility values based on the UK value set recommended by NICE. Patient-level QALYs over 12 months will be estimated using the area under the curve approach. Missing data mechanisms will be assessed, and multiple imputation applied as needed. Imputed datasets will be used for bivariate analyses of costs and QALYs to estimate incremental cost per QALY and associated confidence intervals. Results will be presented on the cost-effectiveness plane, through cost-effectiveness acceptability curves, net monetary benefit calculations, and value of information analyses.

   Limitations of trial-based economic analyses include potential discrepancies between observed and real-world costs, particularly for emerging technologies such as robot-assisted surgery. Costs may vary depending on hospital throughput and market conditions, and sensitivity analyses will explore these variations. If initial 12-month trial data show convergent outcomes and costs or dominance of one surgical approach, analyses will remain within-trial; otherwise, a longer-term economic model incorporating extended trial follow-up and external epidemiological data will be developed.

   Surgical efficiency and task load evaluation An assessment of the surgical efficiency, learning curve, physical and cognitive task load costs associated with VRAS raTKA versus mTKA will be conducted. This will include an evaluation of intraoperative timings, NASA Task Load Index (TLX) scores reported by clinicians based on their perception of surgical cases (after the first and every fifth case, including non-trial training period; i.e. after 1st case, 5th, 10th, 15th etc.), and clinician-reported questionnaires.

   SUMMARY The MARVEL Project is a multicentre, pragmatic, parallel-group, blinded randomised controlled trial evaluating VELYS functional alignment using raTKA against conventional mTKA in primary surgery for osteoarthritis.

   The primary objective is to determine whether raTKA is associated with a greater improvement in patient-reported knee-specific function, measured by change in FJS from baseline to 6 months. Secondary outcomes include additional PROMs (OKS, EQ-5D-5L, pain, satisfaction); functional, physiological, and technology-enabled activity-measures; health service resource use; analgesic use; radiographic alignment; intra-operative digital metrics; surgical efficiency; task load and clinician experience, and safety/adverse events. Gold-standard biomechanical evaluation will be conducted in a motion analysis laboratory setting, assessing gait and validated indicators of knee function, while an integrated health-economic evaluation will assess NHS costs, quality-adjusted life years, and cost-effectiveness.

   An internal pilot will test feasibility with predefined progression criteria. A total of 346 participants will be randomised 1:1 across centres, with a 12-month follow-up for clinical and functional outcomes, and survivorship evaluation for a minimum of 10 years. Results will deliver definitive evidence on the clinical, biomechanical, and health economic impact of imageless robot-assisted functional TKA compared with standard mTKA.

Inclusion Criteria:

• Aged >18 years
• Primary knee OA
• Requiring TKA
• Suitable for either raTKA or mTKA
• Capable to provide written informed consent
• Capable and willing to complete outcome measures throughout study period

Exclusion Criteria:

• Inflammatory arthritis
• Significant symptomatic hip, ankle, or contralateral knee OA
• Prior TKA or major injury or major surgery on the index knee
• Neuromuscular gait disorders
• BMI >45 kg/m2
• Coronal plane deformity >15° varus/valgus on pre-op HKA radiograph
• Fixed flexion deformity greater than 15°
• Bone loss likely to require non-standard implants (e.g. stems, augments)
• Severe deformity or ligament insufficiency likely to require constraint
• Inability to comply with the required follow-up assessments or complete PROMs
• Contraindications to any aspect of the surgical technique

The final statistical analysis plan will be available on reasonable request from the corresponding author, subject to sponsor approval.

Following publication of the primary trial results, a summary of the trial results will be made available through the public trial registry. De-identified individual participant data underlying the reported results, together with the data dictionary and statistical code required to reproduce the primary analysis, will be considered for sharing with bona fide researchers on reasonable request.

Beginning from the date of final end trial results publication

De-identified individual participant data underlying the reported results, together with the data dictionary and statistical code required to reproduce the primary analysis, will be considered for sharing with bona fide researchers on reasonable request.

Requests must include a methodologically sound proposal and may require evidence of appropriate ethical and institutional approvals and completion of a data sharing agreement. Data will only be shared where this is consistent with participant consent, sponsor requirements, data protection legislation, and institutional information governance procedures. Requests should be directed to the study CI.

The study is a randomised controlled trial with patient and outcome assessor blinding. It is not possible to blind surgeons given the nature of the interventions. Non-surgical clinical staff (e.g. ward nurses and allied health professionals) will be blinded. Femoral and tibial tracker pins will be placed within the surgical incision, avoiding the need for accessory incisions in raTKA cases or sham incisions to aid blinding in mTKA cases. Documentation will be redacted appropriately to prevent inadvertent unblinding of clinicians. PROMs will be completed independently by participants. Radiographic assessments and biomechanical analyses will be performed by assessors who are blinded to treatment allocation. All study data will be anonymised and the study arm concealed before the conduct of data analysis. No unblinding procedures are planned.